Autism refers to a group of pervading neurodevelopmental disabilities that are characterized by abnormalities in social, communicative and behavioral functioning. The symptoms of autism are detectable in early childhood, and persist throughout the life span of affected individuals. The prevalence of autism in general population is as high as 1 in 1000, and affects individuals of every race and ethnic background. Most subjects with autism have mental deficiencies, and often develop epileptic seizures. While there are strong evidences in support of underlying genetic etiology, so far, molecular biological defects of autism are unknown. Our preliminary data in three out of four autopsy brain samples from individuals affected with autism, showed complete absence of a 29 kDa protein with a pI of about 5.9 and marked increase of a 36 kDa protein with a pl of about 4.9. The fourth brain lack these abnormalities but showed absence of another protein of molecular weight 22 kDa with a pI of about 7.0. The objectives of the present research proposal are to isolate and characterize these proteins, and determine whether analyses for these aberrant proteins can be used as biological markers in identifying certain types of autism. To reach this goal, we will partially purify these proteins from autopsy human brain samples using classical chromatographic approaches. Amino-terminal or internal sequences of these proteins will be determined, and a sequence comparison with the protein sequence database will reveal whether these are novel proteins. If these are proteins of known functions, then we will develop biochemical assays for their detection in tissues and lymphoid cells. These biological marker will be based either on functions, or levels of these proteins determined by western blot analyses using antibodies developed against synthetic peptides.